MONOCHROMATIC LIGHT OF ANY DESIRED WAVE-LENGTH. 
925 
as the centre of the line is the part adjusted there is no real necessity to stop the slit 
down with one of the smaller stops. If, however, the f-inch or ^-inch stop is placed 
before the receiving slit the lines are then apparently quite straight and fall wholly 
in the central line between the two jaws when adjusted. 
The readings of the prism circle are then taken, with the aid of tL.e vernier, for 
each of the lines so adjusted by suitable movement of the prism, and recorded in a 
table. This table should be supplemented by a curve, in order that the readings for 
intermediate wave-lengths may be obtained by interpolation. The readings for the 
solar lines of hydrogen may be confirmed, if considered desirable, by use of a 
hydrogen Geissler tube illuminated by means of four Grove’s cells and an induction 
coil. It is also convenient to confirm the sodium readings, and to extend the 
observations by recording the readings for the red lithium and the green thallium 
line. For this purpose it is convenient to have at hand a metal-lined box, fitted 
with a wiiidow in front and a door behind, a chimney above and air holes in the 
raised base, containing a Bunsen lamp and an arrangement for bringing one of tliree 
platinum spoons, containing respectively a supply of a sodium salt, a lithium salt, and 
a smaller quantity of a thallium salt, into the flame as desired by means of a rotating 
arrangement manipulated from outside by means of a lever. This arrangement is also 
very convenient for confirming the circle readings before and after every important 
investigation, in order to be quite certain that no movement of any of the parts of the 
instrument shall have occurred. For this purpose it is sufficient to ascertain whether 
the reading for the double sodium line remains the same. It is thus only necessary 
to use the poisonous thallium vapour for a few seconds during the first determination 
of the reading for that wave-length. Although the exit slit frequently rec[uires 
slight opening or closing, to suit the lesser or more perfect transparency of the 
crystals examined, the readings for the sodium and hence for all the lines have never 
been found to vary by more than two minutes of arc. 
After thus determining the relation between the wave-length of the issuing light 
and the prism circle readings, the mirror and the eye-piece are removed from before 
their respective slits, and it is only necessary, when at any time monochromatic light 
of any specific wave-length is required, to set the prism so that the circle reading is 
identical with that recorded in the table, or obtained by interpolation from the curve, 
for light of the wave-length in question, and to illuminate the slit of the collimator 
with a sufficiently powerful artificial source of light. Tho oxy-coal gas lime-light 
afiords ample illumination with slits nearly closed, and if the observer is sufficiently 
fortunate to have an electric arc lamp at his disposal the openings of the slits may be 
so fine that the slightest further movement of the milled head of the tangent screw 
closes them altogether. If the receiving slit is opened to the extent of a third of a 
millimetre and the exit slit to about a quarter of a millimetre, sufficiently good 
illumination may be obtained with wave-lengths up to F by employing the incan¬ 
descent gas-light previously referred to in the lantern instead of the lime or electric 
light, retaining the condensers of the lantern to concentrate the rays upon the slit. 
